Cloud Computing Environment Gml Spatial Data Query And Spatial Analysis

GML (Geography Markup Language) developed by the OGC has been proposed for spatial data’smodeling, integration and sharing and provided unified standards and framework. With GML,there is no loss between different GIS data’s transform. Along with the rapid development of GIStechnology, the problem GIS need to be solved is becoming scale and complex and people’sdemand for spatial information is increasing, GML data is in larger and larger number, singlecomputer storage and management can’t meet these needs. How to efficiently deal with thesemassive, semi-structured GML data become a concern. Cloud computing being concerned canorganize and manage huge amounts of data and provide an effective, intelligent solution.Cloudcomputing has a lot of advantages such as having a very large scale, high reliability, highscalability, low cost, versatility, and on-demand services. How to introduce cloud computing toprocess massive amounts GML data, is initial reason of the study.Aidding Hadoop open source cloud computing platform, together with the study of distributedstorage and parallel index GML being studied, research focus on followings:Firstly, built the GML multi-level parallel R-tree index. Parsed GML by the SAX technology,parted GML spatial data based on Hilbert space filling curve data in GML spatial datapartitioning strategy, built multi-level R-tree indexing mechanism. With the filtering action of theindex space, researched GML parallel query.Secondly, aided Hadoop distributed computing environment to construct GML parallel querymechanism. After filtered by the first spatial index, got some very small candidate sets, and thenran the sub parallel computing queries.Thirdly, aided Hadoop distributed parallel computing environment to finish GML spatial analysisfunction. In order to ensure the load balance of GIS operations, designed parallel task allocationand task scheduling the two load balancing model. Based on Hadoop nature of the localcalculation, assumed GML data had been storted roughly balanced in the system.Fourthly, verified GML query and spatial analysis’ performance under Hadoop. Vetifiedy by thecomparison of the case in single PC.This paper built a theoretical framework of parallel query and parallel analysis of GML spatialdata, achieved some functionality. With the advantages of Hadoop cloud computing platform bothpromoted GML application in depth study, promoted the GML-based spatial data sharingtechnology’s development, also provided a quick, cheap and reliable method to manage andprocess massive amounts of GML spatial data. Meanwhile, using cloud computing applications todeal with spatial data, which also a drive for cloud computing.